1. A coronary artery segmentation method, comprising:

determining first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on coronary artery segmentation information corresponding to the coronary artery image to be segmented by using a trunk branch model;

and determining first coronary artery main trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using a main trunk branch segmentation model.

2. The coronary artery segmentation method according to claim 1, wherein after determining the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented by using the trunk branch model, the method further comprises:

performing a first post-processing operation on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information to obtain second coronary artery trunk information and second coronary artery branch information;

determining, by using a trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information, including:

determining, using the trunk branch segmentation model, the first coronary trunk segmentation information and the first coronary branch segmentation information based on the second coronary trunk information and the second coronary branch information.

3. The coronary artery segmentation method according to claim 2, wherein the performing a first post-processing operation on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information to obtain second coronary artery trunk information and second coronary artery branch information comprises:

determining a coronary segmentation centerline based on the first coronary trunk information and the first coronary branch information;

determining a seed point based on the coronary segmentation centerline, the first coronary trunk information, and the first coronary branch information;

performing region growing based on the seed points to determine the second coronary trunk information and the second coronary branch information.

4. The coronary artery segmentation method according to any one of claims 1 to 3, wherein after determining first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information by using the trunk branch segmentation model, the method further comprises:

and performing second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information.

5. The coronary artery segmentation method according to claim 4, wherein the performing a second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information comprises:

determining a maximum connected domain of a coronary artery trunk based on the first coronary artery trunk segmentation information;

performing region growing based on the maximum connected domain of the coronary artery trunk to determine the second coronary artery trunk segmentation information;

determining a coronary branch connected domain based on the first coronary branch segmentation information;

determining the second coronary branch segmentation information based on the coronary branch connected domain.

6. The coronary artery segmentation method according to any one of claims 1 to 3, wherein the determining, by using the trunk branch model, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented comprises:

and inputting the aorta segmentation information, the heart segmentation information and the coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model to obtain the first coronary artery trunk information and the first coronary artery branch information.

7. The coronary artery segmentation method according to claim 6, wherein before the inputting the aorta segmentation information, the heart segmentation information and the coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model to obtain the first coronary artery trunk information and the first coronary artery branch information, further comprising:

determining a training data set and aorta segmentation information, heart segmentation information and coronary artery segmentation information corresponding to the training data set;

determining external coronary artery graphic information corresponding to the coronary artery segmentation information corresponding to the training data set;

sequentially determining external aorta graphic information, external heart graphic information and external coronary graphic information based on the external coronary graphic information;

establishing an initial network model, and training the initial network model based on the training data set, the aorta external graph information, the heart external graph information and the coronary artery external graph information to generate the trunk branch model.

8. The coronary artery segmentation method according to any one of claims 1 to 3, wherein the determining, by using the trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information comprises:

inputting the aorta segmentation information, the heart segmentation information, the coronary artery segmentation information, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented into the trunk branch segmentation model to obtain the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information.

9. A coronary artery segmentation device, comprising:

the first determining module is configured to determine first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on coronary artery segmentation information corresponding to the coronary artery image to be segmented by using a trunk branch model; and

and the second determining module is configured to determine first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information by using a trunk branch segmentation model.

10. A computer-readable storage medium, characterized in that the storage medium stores a computer program for performing the coronary segmentation method according to any one of the preceding claims 1 to 8.

11. An electronic device, characterized in that the electronic device comprises:

a processor;

a memory for storing the processor-executable instructions;

the processor configured to perform the coronary segmentation method of any of the preceding claims 1 to 8.

Background

As is well known, the coronary artery (also called coronary artery) is an artery supplying blood to the heart and has a very complicated structure. For example, according to the American Heart Association's revised coronary segmentation method, it is divided into 15 segments in total.

However, the existing coronary artery segmentation method is limited by the complicated structure of the coronary artery, and has the problems of low robustness and poor segmentation effect.

Disclosure of Invention

The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides a coronary artery segmentation method and device, a computer-readable storage medium and electronic equipment.

In a first aspect, an embodiment of the present application provides a coronary artery segmentation method, including: determining first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented by using a trunk branch model; and determining first coronary artery main trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using the main trunk branch segmentation model.

With reference to the first aspect, in certain implementations of the first aspect, after determining, by using the trunk branch model, first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented, the method further includes: performing a first post-processing operation on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information to obtain second coronary artery trunk information and second coronary artery branch information; the method for determining the first coronary artery main trunk segmentation information and the first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using the main trunk branch segmentation model comprises the following steps: determining first coronary artery trunk segmentation information and first coronary artery branch segmentation information based on the second coronary artery trunk information and the second coronary artery branch information by using the trunk branch segmentation model.

With reference to the first aspect, in certain implementations of the first aspect, performing a first post-processing operation on the first coronary trunk information and the first coronary branch information to correct the first coronary trunk information and the first coronary branch information to obtain second coronary trunk information and second coronary branch information includes: determining a coronary artery segmentation centerline based on the first coronary artery trunk information and the first coronary artery branch information; determining a seed point based on the coronary artery segmentation central line, the first coronary artery trunk information and the first coronary artery branch information; region growing is performed based on the seed points to determine second coronary trunk information and second coronary branch information.

With reference to the first aspect, in certain implementations of the first aspect, after determining, by using the trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information, the method further includes: and performing second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information.

With reference to the first aspect, in certain implementations of the first aspect, performing a second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information includes: determining a maximum connected domain of the coronary artery trunk based on the first coronary artery trunk segmentation information; performing region growth based on the maximum connected domain of the coronary artery trunk to determine second coronary artery trunk segmentation information; determining a coronary branch connected domain based on the first coronary branch segmentation information; determining second coronary branch segmentation information based on the coronary branch connected domain.

With reference to the first aspect, in certain implementations of the first aspect, determining, by using a trunk branch model, first coronary trunk information and first coronary branch information corresponding to a coronary image to be segmented based on coronary segmentation information corresponding to the coronary image to be segmented includes: and inputting the aorta segmentation information, the heart segmentation information and the coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model to obtain first coronary artery trunk information and first coronary artery branch information.

With reference to the first aspect, in certain implementations of the first aspect, before inputting the aorta segmentation information, the heart segmentation information, and the coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model to obtain the first coronary artery trunk information and the first coronary artery branch information, the method further includes: determining a training data set and aorta segmentation information, heart segmentation information and coronary artery segmentation information corresponding to the training data set; determining external coronary artery graphic information corresponding to the coronary artery segmentation information corresponding to the training data set; sequentially determining external aorta graphic information, external heart graphic information and external coronary graphic information based on the external coronary graphic information; establishing an initial network model, and training the initial network model based on a training data set, aorta external graph information, heart external graph information and coronary artery external graph information to generate a trunk branch model.

With reference to the first aspect, in certain implementations of the first aspect, determining, by using a trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to a coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information includes: and inputting the aorta segmentation information, the heart segmentation information, the coronary artery segmentation information, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented into a trunk branch segmentation model to obtain first coronary artery trunk segmentation information and first coronary artery branch segmentation information.

In a second aspect, an embodiment of the present application provides a coronary artery segmentation apparatus, including: the first determining module is configured to determine first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented by using the trunk branch model; and the second determining module is configured to determine first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information by using the trunk branch segmentation model.

In a third aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program for executing the coronary artery segmentation method mentioned in any one of the above embodiments.

In a fourth aspect, an embodiment of the present application provides an electronic device, including: a processor; a memory for storing processor-executable instructions; the processor is configured to perform the coronary segmentation method mentioned in any of the above embodiments.

According to the coronary artery segmentation method provided by the embodiment of the application, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented are determined by using the trunk branch model; then, determining first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information by using a trunk branch segmentation model; through the effective cooperation of the main branch model and the main branch segmentation model, the problem of poor coronary artery segmentation effect is solved, the robustness of the coronary artery segmentation process is effectively improved, the two models respectively play their own roles, and the corresponding models can be more effectively optimized aiming at specific problems in the actual optimization process, so that the accuracy of the finally determined segmentation result is further improved.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.

Fig. 1 is a schematic view of a scenario applicable to the embodiment of the present application.

Fig. 2 is a schematic view of another scenario applicable to the embodiment of the present application.

Fig. 3 is a schematic flow chart of a coronary artery segmentation method according to an exemplary embodiment of the present application.

Fig. 4 is a schematic flowchart illustrating a process of determining first coronary artery trunk segment information and first coronary artery branch segment information according to another exemplary embodiment of the present application.

Fig. 5 is a flowchart illustrating a first post-processing operation according to an exemplary embodiment of the present application.

Fig. 6 is a schematic flow chart illustrating obtaining second coronary artery trunk segment information and second coronary artery branch segment information according to an exemplary embodiment of the present application.

Fig. 7 is a flowchart illustrating a second post-processing operation according to an exemplary embodiment of the present application.

Fig. 8 is a schematic flow chart illustrating a process of determining first coronary artery trunk information and first coronary artery branch information corresponding to a coronary artery image to be segmented according to an exemplary embodiment of the present application.

Fig. 9 is a schematic flowchart illustrating a process of generating a trunk branch model according to an exemplary embodiment of the present application.

Fig. 10 is a schematic diagram illustrating tag conversion provided by an exemplary embodiment of the present application.

Fig. 11 is a flowchart illustrating a method for training a trunk branch model according to an exemplary embodiment of the present disclosure.

Fig. 12 is a schematic flowchart for determining first coronary artery trunk segment information and first coronary artery branch segment information according to still another exemplary embodiment of the present application.

Fig. 13 is a flowchart illustrating a method for training a trunk branch segmentation model according to an exemplary embodiment of the present application.

Fig. 14 is a schematic flow chart of a coronary artery segmentation method according to still another exemplary embodiment of the present application.

Fig. 15 is a schematic structural diagram of a coronary artery segmentation apparatus according to an exemplary embodiment of the present application.

Fig. 16 is a schematic structural diagram of a coronary artery segmentation apparatus according to another exemplary embodiment of the present application.

Fig. 17 is a schematic structural diagram of a first post-processing operation module according to an exemplary embodiment of the present application.

Fig. 18 is a schematic structural diagram of a coronary artery segmentation apparatus according to another exemplary embodiment of the present application.

Fig. 19 is a schematic structural diagram of a second post-processing operation module according to an exemplary embodiment of the present application.

Fig. 20 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application.

Fig. 21 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application.

Fig. 22 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application.

Fig. 23 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

Summary of the application

The heart is shaped like an inverted, slightly flattened cone, and if it is considered as a head, the coronary artery located at the top of the head, almost encircling the heart, is just like a crown, which is the name of the coronary artery. The coronary artery is the artery supplying blood to the heart, originates in the aortic sinus at the root of the aorta, divides into two branches, and runs on the surface of the heart. Using the existing classification principle, the distribution of coronary arteries is mainly classified into three types, including right dominant, equilibrium, and left dominant. According to the american heart association revised coronary artery segmentation method, the coronary artery is divided into a total of 15 segments, respectively: the proximal segment RCA1 of the right coronary artery, the middle segment RCA2 of the right coronary artery, the distal segment RCA3 of the right coronary artery, the posterior descending branch, the posterior left ventricle, the main trunk of the left coronary artery, the proximal segment of the anterior descending branch, the middle segment of the anterior descending branch, the distal segment of the anterior descending branch, the first diagonal branch, the second diagonal branch, the middle branch, the proximal segment of the circumgyrating branch, the first blunt edge branch, the distal segment of the circumgyrating branch and the second blunt edge branch.

The existing coronary artery segmentation technology mainly comprises two categories of traditional algorithms and deep learning algorithms, and the two categories of algorithms are carried out based on the result of coronary artery segmentation. The traditional algorithm combines the characteristics of the coronary artery segmentation, and establishes a spherical coordinate system or a tree-shaped data structure by utilizing the anatomical shape of each branch segment of the coronary artery and the geometric structural relationship between the branch segments so as to determine the category of each segment. However, the method is limited by the complicated structure of the coronary artery and the effect of the anterior model coronary artery segmentation, and the problem of large error often occurs in the practical application process. The deep learning algorithm utilizes a deep learning semantic segmentation network to perform coronary segmentation to determine the category of each segment. However, the deep learning algorithm is limited by factors such as a complex structure of coronary artery and various labels, and has a problem of poor segmentation effect.

In order to solve the above problems, in order to improve the accuracy of coronary artery segmentation, the embodiment of the present application provides a coronary artery segmentation method, in which a first coronary artery trunk information and a first coronary artery branch information corresponding to a coronary artery image to be segmented are determined based on coronary artery segmentation information corresponding to the coronary artery image to be segmented through a trunk branch model, so as to provide an accurate basis for subsequent coronary artery segmentation; and then, determining first coronary artery main trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using a main trunk branch segmentation model to obtain a final coronary artery segmentation result, thereby determining the category of each segment.

Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.

Exemplary System

Fig. 1 is a schematic view of a scenario applicable to the embodiment of the present application. As shown in fig. 1, a scenario to which the embodiment of the present application is applied includes a server 1 and an image capturing device 2, where there is a communication connection relationship between the server 1 and the image capturing device 2.

In particular, the image acquisition device 2 is used to acquire a coronary image to be segmented. The image acquisition device 1 may be a CT scanner, an X-ray machine, an mri (magnetic Resonance imaging) device, or other devices having an image acquisition function, as long as it can acquire a coronary image, and the structure of the image acquisition device 2 is not specifically limited in the present application.

The server 1 may be one server, a server group composed of a plurality of servers, or a virtualization platform or a cloud computing service center, and the type of the server 1 is not specifically limited in the present application. The server 1 determines first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented acquired by the image acquisition device 2 by using the trunk branch model, and then determines first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information by using the trunk branch segmentation model. That is, this scenario implements a coronary segmentation method.

Since the scene shown in fig. 1 uses the server 1 to implement the coronary artery segmentation method, the scene not only can improve the adaptability of the scene, but also can effectively reduce the calculation amount of the image acquisition device 2.

It should be noted that the present application is also applicable to another scenario. Fig. 2 is a schematic view of another scenario applicable to the embodiment of the present application. Specifically, the scene includes an image processing device 3, wherein the image processing device 3 includes an image acquisition module 301 and a calculation module 302, and a communication connection relationship exists between the image acquisition module 301 and the calculation module 302.

Specifically, the image acquisition module 301 in the image processing apparatus 3 is configured to acquire a coronary artery image to be segmented, the calculation module 302 in the image processing apparatus 3 determines, by using the trunk branch model, first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented acquired by the image acquisition module 301, and then determines, by using the trunk branch segmentation model, the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information. That is, this scenario implements a coronary segmentation method.

Since the above-mentioned scenario shown in fig. 2 implements the coronary artery segmentation method by using the image processing apparatus 3, and does not need to perform data transmission operations with a server or other related devices, the above-mentioned scenario shown in fig. 2 can ensure the real-time performance of the coronary artery segmentation method.

Exemplary method

Fig. 3 is a schematic flow chart of a coronary artery segmentation method according to an exemplary embodiment of the present application. As shown in fig. 3, the coronary artery segmentation method provided by the embodiment of the present application includes the following steps.

And step 10, determining first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented by using the trunk branch model.

The coronary image to be segmented mentioned in step 10 may be, for example, a coronary image of the subject to be evaluated taken by a general-purpose or special-purpose photographing apparatus.

The coronary artery image to be segmented may be a Computed Tomography (CT) image, a Magnetic Resonance Imaging (MRI) image, a Computed Radiography (CR) image, or a Digital Radiography (DR) image, which is not specifically limited in this embodiment of the present invention. The coronary artery segmentation method provided by the embodiment of the application can be applied to all coronary artery images and has universality.

The embodiment of the present application does not limit the specific form of the medical image to be segmented, and may be an original medical image, a preprocessed medical image, or a partial image series in the original medical image, that is, a part of the original medical image. In addition, the acquisition object corresponding to the medical image to be processed can be a human body or an animal body.

Illustratively, the coronary segmentation information corresponding to the coronary image to be segmented mentioned in step 10 is a coronary segmentation mask image obtained based on the pre-model. For example, the coronary segmentation mask image is a binary image composed of 0 and 1, where the 0 value region is the background and the 1 value region is the coronary. Whether the pixel points in the image belong to the coronary artery region or the background region irrelevant to the coronary artery region can be distinguished through two values of 0 and 1.

Illustratively, the first coronary artery trunk information mentioned in step 10 corresponds to a first coronary artery trunk segmentation mask image, and the first coronary artery branch information corresponds to a first coronary artery branch segmentation mask image. The main branch model is a semantic segmentation model obtained through model training and is used for outputting a first coronary artery main segmentation mask image and a first coronary artery branch segmentation mask image corresponding to the coronary artery image to be segmented. Considering that the trunk and the branch of the coronary artery are easy to be divided, the trunk area and the branch area are divided firstly, and an accurate foundation is provided for subsequent coronary artery segmentation.

And step 20, determining first coronary artery main trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using the main trunk branch segmentation model.

Illustratively, the first coronary artery trunk segmentation information mentioned in step 20 corresponds to a coronary artery trunk segmentation mask image; the first coronary branch segmentation information corresponds to a coronary branch segmentation mask image. The main trunk branch segmentation model is a semantic segmentation model obtained through model training and is used for outputting a coronary artery main trunk segmentation mask image and a coronary artery branch segmentation mask image corresponding to the coronary artery image to be segmented.

It should be noted that, in the embodiment of the present application, specific network structures of the trunk branch model and the trunk branch segmentation model are not limited, and the trunk branch model and the trunk branch segmentation model may be formed by any type of neural network. Alternatively, the Neural Network may be a Convolutional Neural Network (CNN), a Deep Neural Network (DNN), a Recurrent Neural Network (RNN), or the like. This is not particularly limited in the embodiments of the present application.

According to the coronary artery segmentation method provided by the embodiment of the application, firstly, a main trunk branch model is utilized, first coronary artery main trunk information and first coronary artery branch information corresponding to a coronary artery image to be segmented are determined based on coronary artery segmentation information corresponding to the coronary artery image to be segmented, and an accurate basis is provided for subsequent coronary artery segmentation; and then, determining first coronary artery main trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery main trunk information and the first coronary artery branch information by using a main trunk branch segmentation model to obtain a final coronary artery segmentation result, thereby determining the category of each segment. In addition, the coronary artery segmentation method provided by the application realizes the coronary artery segmentation by establishing the serially connected double-model semantic segmentation network, namely a large segmentation module comprises two models which are mutually serially connected, and the second model fully utilizes the output result of the first model for prediction, so that the task load of the second module is greatly reduced. The two models are effectively matched, so that the problem of poor coronary artery segmentation effect is solved, the robustness of the coronary artery segmentation process is effectively improved, the two models respectively play their own roles, the corresponding models can be more effectively optimized aiming at specific problems in the actual optimization process, and the accuracy of the finally determined segmentation result is further improved.

Fig. 4 is a schematic flowchart illustrating a process of determining first coronary artery trunk segment information and first coronary artery branch segment information according to another exemplary embodiment of the present application. The embodiment shown in fig. 4 of the present application is extended based on the embodiment shown in fig. 3 of the present application, and the differences between the embodiment shown in fig. 4 and the embodiment shown in fig. 3 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 4, in the coronary artery segmentation method provided in the embodiment of the present application, after determining, by using the trunk branch model, first coronary artery trunk information and first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented, the following steps are further included.

And step 30, performing a first post-processing operation on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information to obtain second coronary artery trunk information and second coronary artery branch information.

Specifically, if one model is used for division when segmenting the coronary artery, the requirement on the model is high, and when the model is wrong, the model is difficult to correct by using post-processing. Aiming at the structural characteristics of coronary artery complexity, the two models which are connected in series are adopted in the embodiment of the application, and each model can independently correct the prediction error of each model by using post-processing operation. The first post-processing operation mentioned in step 30 is used to correct the prediction result output by the trunk branch model, and filter the abnormal value in the prediction result, so as to obtain more accurate second coronary artery trunk information and second coronary artery branch information.

In the embodiment of the present application, determining, by using a trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to a coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information includes the following steps.

And step 40, determining first coronary artery trunk segmentation information and first coronary artery branch segmentation information based on the second coronary artery trunk information and the second coronary artery branch information by using the trunk branch segmentation model.

Specifically, after the first coronary artery trunk information and the first coronary artery branch information are subjected to the first post-processing operation, abnormal values and error information in the first coronary artery trunk information and the first coronary artery branch information are filtered and corrected, and the obtained second coronary artery trunk information and second coronary artery branch information are effective information. Therefore, the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information obtained based on the effective second coronary artery trunk information and the effective second coronary artery branch information are more accurate by utilizing the trunk branch segmentation model.

According to the coronary artery segmentation method provided by the embodiment of the application, the first post-processing operation is carried out on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information to obtain the second coronary artery trunk information and the second coronary artery branch information, then the trunk branch segmentation model is utilized to determine the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information based on the second coronary artery trunk information and the second coronary artery branch information, and the problem that the result of coronary artery segmentation is inaccurate due to the existence of abnormal values is solved. In addition, the trunk branch model corrects the output result by using a first post-processing operation, and the corrected second coronary artery trunk information and the corrected second coronary artery branch information are used as the input of the trunk branch segmentation model, so that the prediction result of the trunk branch segmentation model is more accurate, and the situation of low robustness of the coronary artery segmentation caused by inaccurate segmentation is further reduced.

Fig. 5 is a flowchart illustrating a first post-processing operation according to an exemplary embodiment of the present application. The embodiment shown in fig. 5 of the present application is extended on the basis of the embodiment shown in fig. 4 of the present application, and the differences between the embodiment shown in fig. 5 and the embodiment shown in fig. 4 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 5, in the coronary artery segmentation method provided in the embodiment of the present application, a first post-processing operation is performed on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information, so as to obtain second coronary artery trunk information and second coronary artery branch information, including the following steps.

Step 301, determining a coronary artery segmentation centerline based on the first coronary artery trunk information and the first coronary artery branch information.

Specifically, the first coronary artery trunk information includes coronary artery trunk segmentation data, and the first coronary artery branch information includes coronary artery branch segmentation data. And performing centerline extraction operation based on the coronary artery trunk segmentation data and the coronary artery branch segmentation data to obtain a coronary artery segmentation centerline.

Illustratively, the operation of extracting the central line can be extracted based on the lee94 algorithm, and can also be extracted based on a skeletonization algorithm. As long as the coronary artery segmentation centerline can be determined based on the first coronary artery trunk information and the first coronary artery branch information, the specific algorithm for extracting the centerline is not limited in the embodiment of the present application.

Step 302, determining a seed point based on the coronary artery segmentation centerline, the first coronary artery trunk information and the first coronary artery branch information.

Specifically, the coronary artery branches from the trunk structure one level by one level to form the complicated multi-level branches, and based on the characteristics of the multi-level branch structure of the coronary artery, the complicated coronary artery structure can be further classified in a refined manner by utilizing the bifurcation points and the end points of the branch structure. Wherein, the coronary segmentation midline can be understood as a series of continuous points on the center of the coronary segmentation mask image. For points on the coronary segmentation midline, the coronary segmentation midline is divided into line segments, defined herein as a series of points from end point to end point or end point to bifurcation point, depending on whether it is a bifurcation point and an end point. All points on the line segment constitute a set of line segment points. For each line segment, each point in the line segment point set corresponds to a prediction result output by the trunk branch model, whether each point is first coronary artery trunk information or first coronary artery branch information is determined according to the mode of the prediction result of the trunk branch model, and finally the line segment point result in the line segment point set is used as a seed point.

Step 303, performing region growing based on the seed points to determine second coronary artery trunk information and second coronary artery branch information.

Specifically, the seed point in step 302 is used as a starting point of growth, and the coronary artery segmentation mask image corresponding to the first coronary artery trunk information and the first coronary artery branch information is used as an orbit region, and a region growing operation is performed, so that a more accurate trunk branch result is obtained.

According to the coronary artery segmentation method provided by the embodiment of the application, the central line of the coronary artery segmentation is extracted, then the result of the line segment point determined by the central line of the coronary artery segmentation is used as the seed point, and the region growth is carried out in the mask image of the coronary artery segmentation to determine the second coronary artery trunk information and the second coronary artery branch information. The region growing is carried out based on the result predicted by the trunk branch model, so that the obtained trunk branch result is more accurate, and an accurate basis is provided for subsequent coronary artery segmentation.

Fig. 6 is a schematic flow chart illustrating obtaining second coronary artery trunk segment information and second coronary artery branch segment information according to an exemplary embodiment of the present application. The embodiment shown in fig. 6 of the present application is extended based on the embodiment shown in fig. 3 of the present application, and the differences between the embodiment shown in fig. 6 and the embodiment shown in fig. 3 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 6, in the coronary artery segmentation method provided in the embodiment of the present application, after determining, by using the trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information, the following steps are further included.

And step 50, performing second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information.

Specifically, the second coronary artery trunk segment information and the second coronary artery branch segment information are effective information obtained through a second post-processing operation, which has filtered and corrected abnormal values and error information in the first coronary artery trunk segment information and the first coronary artery branch segment information.

According to the coronary artery segmentation method provided by the embodiment of the application, the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information are corrected through second post-processing operation, so that the second coronary artery trunk segmentation information and the second coronary artery branch segmentation information are obtained, and the purpose of improving the accuracy of the coronary artery segmentation is achieved. In addition, the result output by the trunk branch segmentation model is corrected through the second post-processing operation, so that the corrected second coronary artery trunk segmentation information and the corrected second coronary artery branch segmentation information are more accurate, and the situation that the coronary artery segmentation is low in robustness due to inaccurate segmentation is further reduced.

Fig. 7 is a flowchart illustrating a second post-processing operation according to an exemplary embodiment of the present application. The embodiment shown in fig. 7 of the present application is extended based on the embodiment shown in fig. 6 of the present application, and the differences between the embodiment shown in fig. 7 and the embodiment shown in fig. 6 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 7, in the coronary artery segmentation method provided in the embodiment of the present application, a second post-processing operation is performed on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information, so as to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information, including the following steps.

Step 501, determining a maximum connected domain of a coronary artery trunk based on first coronary artery trunk segmentation information.

Specifically, the first coronary artery trunk segmentation information corresponds to a coronary artery trunk segmentation mask image, each trunk in the coronary artery trunk segmentation mask image corresponds to a category predicted by a trunk branch segmentation model, a maximum connected domain corresponding to each trunk category is reserved, and other unreserved trunk regions grow along the maximum connected domain in a region growing mode to obtain the categories of other trunk regions.

It should be noted that the coronary artery trunk segmented mask image corresponding to the first coronary artery trunk segmented information may be subjected to gray processing, a gray value is obtained, then a region with a gray value larger than a preset gray value is extracted, and a region with a largest area in the region with a gray value larger than the preset gray value is selected as a largest connected domain.

And 502, performing region growth based on the maximum connected domain of the coronary artery trunk to determine second coronary artery trunk segmentation information.

Specifically, the region growing is performed by taking the maximum connected domain of the coronary artery trunk as the center, and the second coronary artery trunk segmentation information is obtained. The method is simple and high in efficiency, and more accurate second coronary artery trunk segmentation information is obtained in a mode of carrying out region growing by taking the maximum connected domain as the center.

Step 503, determining a coronary branch connected domain based on the first coronary branch segmentation information.

Specifically, the first coronary branch segment information corresponds to a coronary branch segment mask image. And for the coronary branch segmented mask image, solving each coronary branch connected domain.

Step 504, second coronary branch segmentation information is determined based on the coronary branch connected domain.

Specifically, each coronary branch connected domain corresponds to one branch type result output by the trunk branch segmentation model, the mode of the type of each coronary branch connected domain is determined based on the branch type result, and the mode of the type is used as the type of the coronary branch connected domain, so that more accurate second coronary branch segmentation information is determined.

The coronary artery segmentation method provided by the embodiment of the application realizes the purpose of correcting the result output by the main branch segmentation model by a mode of determining the second coronary artery main trunk segmentation information and the second coronary artery branch segmentation information based on the maximum connected domain of the coronary artery main trunk and the connected domain of the coronary artery branches. Because each trunk branch forms an independent connected domain, the categories of the trunks and the branches can be further divided by means of the properties of the connected domains, more accurate second coronary artery trunk segmentation information and second coronary artery branch segmentation information are obtained, and the precision and the robustness of the coronary artery segmentation are improved.

Fig. 8 is a schematic flow chart illustrating a process of determining first coronary artery trunk information and first coronary artery branch information corresponding to a coronary artery image to be segmented according to an exemplary embodiment of the present application. The embodiment shown in fig. 8 of the present application is extended based on the embodiment shown in fig. 3 of the present application, and the differences between the embodiment shown in fig. 8 and the embodiment shown in fig. 3 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 8, in the coronary artery segmentation method provided in the embodiment of the present application, the determining, by using the trunk branch model, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented based on the coronary artery segmentation information corresponding to the coronary artery image to be segmented includes the following steps.

And 104, inputting the aorta segmentation information, the heart segmentation information and the coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model to obtain first coronary artery trunk information and first coronary artery branch information.

Specifically, the aorta segmentation information corresponds to an aorta segmentation mask image, the heart segmentation information corresponds to a heart segmentation mask image, and the coronary artery segmentation information corresponds to a coronary artery segmentation mask image. The input information of the main branch model comprises a heart segmentation mask image and an aorta segmentation mask image at the same time, and compared with the method of directly segmenting based on the coronary artery segmentation mask image, the method provides auxiliary feature information for coronary artery segmentation.

According to the coronary artery segmentation method provided by the embodiment of the application, in the process of segmenting the main trunk and the branches based on the coronary artery segmentation information, the heart segmentation information, the aorta segmentation information and other auxiliary characteristics are introduced, so that the structural information and the blood supply position information of the coronary artery are kept as far as possible, the difficulty of subsequent coronary artery segmentation is reduced, and the precision of the coronary artery segmentation is improved.

Fig. 9 is a schematic flowchart illustrating a process of generating a trunk branch model according to an exemplary embodiment of the present application. The embodiment shown in fig. 9 of the present application is extended based on the embodiment shown in fig. 8 of the present application, and the differences between the embodiment shown in fig. 9 and the embodiment shown in fig. 8 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 9, in the coronary artery segmentation method provided in the embodiment of the present application, before the aorta segmentation information, the heart segmentation information, and the coronary artery segmentation information corresponding to the coronary artery image to be segmented are input into the trunk branch model to obtain the first coronary artery trunk information and the first coronary artery branch information, the following steps are further included.

Step 100, determining a training data set and aorta segmentation information, heart segmentation information and coronary artery segmentation information corresponding to the training data set.

Illustratively, the training data set referred to in step 100 contains input samples and annotation target samples. The input samples are heart segmentation mask images, aorta segmentation mask images and coronary artery segmentation mask images, and the target samples are marked as corresponding coronary artery segmentation mask images. When a training data set is obtained, label conversion needs to be performed on a training sample image, so that a mask image corresponding to the training data set is obtained.

Step 101, determining external coronary artery graphic information corresponding to the coronary artery segmentation information corresponding to the training data set.

Specifically, the coronary artery segmentation mask image corresponding to the coronary artery segmentation information comprises a background region and a coronary artery region, and the maximum circumscribed cube region of the coronary artery segmentation mask image is obtained, so that the coronary artery region is extracted, the background region is filtered, a better segmentation effect can be realized, and an important effect is played on subsequent segmentation of the interested feature.

And 102, sequentially determining external aorta graphic information, external heart graphic information and external coronary graphic information based on the external coronary graphic information.

Specifically, a coronary artery segmentation mask image, a heart segmentation mask image and an aorta segmentation mask image are sequentially captured according to the maximum circumscribed cube. It is understood that the coronary artery rectangular frame is cut out at the same position as the cut-out region of the coronary artery segmentation mask image on the heart segmentation mask image and the aorta segmentation mask image, respectively.

Step 103, establishing an initial network model, and training the initial network model based on the training data set, the aorta external graph information, the heart external graph information and the coronary artery external graph information to generate a trunk branch model.

Specifically, the initial network model is a semantic segmentation network model, and the aorta external graph information, the heart external graph information and the coronary artery external graph information are input into the initial network model and trained to generate a trunk branch model.

According to the coronary artery segmentation method provided by the embodiment of the application, the trunk branch model is generated through model training, the trunk branch model is segmented based on the coronary artery image to be segmented to obtain the first coronary artery trunk information and the first coronary artery branch information, the segmentation results of the trunk region and the branch region are optimized, and an accurate basis is provided for subsequent coronary artery segmentation.

Fig. 10 is a schematic diagram illustrating tag conversion provided by an exemplary embodiment of the present application. As shown in fig. 10, a represents 15 stems and branches with different color depths, and b represents a coronary artery division mask image after label conversion, wherein the coronary artery division mask image only has two colors of dark color and light color, the dark color represents the stems, and the light color represents the branches.

Fig. 11 is a flowchart illustrating a method for training a trunk branch model according to an exemplary embodiment of the present disclosure. As shown in fig. 11, an initial network model is first established, a coronary artery segmentation mask image, a heart segmentation mask image and an aorta segmentation mask image are determined, the heart segmentation mask image and the aorta segmentation mask image are sequentially intercepted according to a maximum circumscribed cube determined by the coronary artery segmentation mask image, and the intercepted result is input into the initial network model. Meanwhile, the coronary artery segmented mask images are subjected to gathering operation, two classification mask images of a trunk and branches are obtained, and the two classification mask images are used as a target set. And then constructing a loss function, wherein the loss function is calculated based on the output result of the initial network model and the target set, and comprises a cross entropy loss function or a focal loss function. Finally, training and learning the initial network model by using the training data set in an iterative manner, so that the loss function value tends to be the minimum value; judging whether the model is converged, if so, outputting the model to generate a trunk branch model, otherwise, continuing the training process.

Fig. 12 is a schematic flowchart for determining first coronary artery trunk segment information and first coronary artery branch segment information according to still another exemplary embodiment of the present application. The embodiment shown in fig. 12 of the present application is extended based on the embodiment shown in fig. 3 of the present application, and the differences between the embodiment shown in fig. 12 and the embodiment shown in fig. 3 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 12, in the coronary artery segmentation method provided in the embodiment of the present application, determining, by using a trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to a coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information includes the following steps.

Step 201, inputting aorta segmentation information, heart segmentation information, coronary artery segmentation information, first coronary artery trunk information and first coronary artery branch information corresponding to a coronary artery image to be segmented into a trunk branch segmentation model to obtain first coronary artery trunk segmentation information and first coronary artery branch segmentation information.

Specifically, the first coronary artery trunk information corresponds to a first coronary artery trunk segmentation mask image, and the first coronary artery branch information corresponds to a first coronary artery branch segmentation mask image. Compared with the method for segmenting the coronary artery directly according to the coronary artery segmentation information, the input information of the main branch segmentation model simultaneously comprises the prediction information of the main branch model and the heart and aorta auxiliary characteristic information, the task of the main branch segmentation model is greatly reduced, and the coronary artery segmentation is simpler.

According to the coronary artery segmentation method provided by the embodiment of the application, the aorta segmentation information, the heart segmentation information, the coronary artery segmentation information, the first coronary artery trunk information and the first coronary artery branch information corresponding to the coronary artery image to be segmented are input into the trunk branch segmentation model, the input of the second model comprises the prediction information of the first model, so that the prediction is simpler, and the obtained first coronary artery trunk segmentation information and the obtained first coronary artery branch segmentation information are more accurate.

Fig. 13 is a flowchart illustrating a method for training a trunk branch segmentation model according to an exemplary embodiment of the present application. As shown in fig. 13, the training flow of the trunk branch segmentation model is substantially similar to that of the trunk branch model. Firstly, an initial network segmentation model is established, the input of the initial network segmentation model is an aorta segmentation mask image, a heart segmentation mask image, a coronary artery segmentation mask image, a first coronary artery trunk segmentation mask image and a first coronary artery branch segmentation mask image, external cubic areas of the aorta segmentation mask image, the heart segmentation mask image, the first coronary artery trunk segmentation mask image and the first coronary artery branch segmentation mask image are respectively and sequentially intercepted according to the coronary artery segmentation mask image and are input to the initial network segmentation model, and the coronary artery segmentation mask image is used as a target set. And constructing a loss function, calculating the loss function based on the output result of the initial network segmentation model and the target set, and outputting the model when the model is converged to generate a trunk branch model.

Fig. 14 is a schematic flow chart of a coronary artery segmentation method according to still another exemplary embodiment of the present application. As shown in fig. 14, a coronary artery segmentation mask image, a heart segmentation mask image, and an aorta segmentation mask image are determined first, the heart segmentation mask image and the aorta segmentation mask image are respectively sequentially cut out according to the largest circumscribed cube region of the coronary artery segmentation mask image, and the cut-out region images are input as a trunk branch model. And predicting through the trunk branch model to obtain a prediction result of the trunk branch model. And (3) obtaining a main branch mask image by post-processing the prediction result of the main branch model, and inputting the main branch mask image, the coronary artery segmentation mask image, the heart segmentation mask image and the aorta segmentation mask image into the main branch segmentation model to obtain a main branch classification result. And carrying out post-processing operation on the prediction result of the trunk branch segmentation model, correcting the error of model prediction, and obtaining the final coronary artery segmentation result.

Fig. 15 is a schematic structural diagram of a coronary artery segmentation apparatus according to an exemplary embodiment of the present application. As shown in fig. 15, a coronary artery segmentation apparatus provided in an embodiment of the present application includes:

a first determining module 100 configured to determine, by using a trunk branch model, first coronary artery trunk information and first coronary artery branch information corresponding to a coronary artery image to be segmented based on coronary artery segmentation information corresponding to the coronary artery image to be segmented; and

the second determining module 200 is configured to determine, by using the trunk branch segmentation model, first coronary artery trunk segmentation information and first coronary artery branch segmentation information corresponding to the coronary artery image to be segmented based on the first coronary artery trunk information and the first coronary artery branch information.

Fig. 16 is a schematic structural diagram of a coronary artery segmentation apparatus according to another exemplary embodiment of the present application. The embodiment shown in fig. 16 is extended from the embodiment shown in fig. 15 of the present application, and the differences between the embodiment shown in fig. 16 and the embodiment shown in fig. 15 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 16, the coronary artery segmentation apparatus provided in the embodiment of the present application further includes:

the first post-processing operation module 300 is configured to perform a first post-processing operation on the first coronary artery trunk information and the first coronary artery branch information to correct the first coronary artery trunk information and the first coronary artery branch information, so as to obtain second coronary artery trunk information and second coronary artery branch information.

Wherein the second determining module 200 comprises: determining first coronary artery trunk segmentation information and first coronary artery branch segmentation information based on the second coronary artery trunk information and the second coronary artery branch information by using the trunk branch segmentation model.

Fig. 17 is a schematic structural diagram of a first post-processing operation module according to an exemplary embodiment of the present application. The embodiment shown in fig. 17 is extended from the embodiment shown in fig. 16 of the present application, and the differences between the embodiment shown in fig. 17 and the embodiment shown in fig. 16 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 17, in the coronary artery segmentation apparatus provided in the embodiment of the present application, the first post-processing operation module includes:

a coronary-artery-division-neutral-line determination unit 3001 configured to determine a coronary-artery division neutral line based on the first coronary-artery trunk information and the first coronary-artery branch information;

a seed point determination unit 3002 for determining a seed point based on the coronary artery segmentation centerline, the first coronary artery trunk information, and the first coronary artery branch information;

a second coronary artery trunk and branch information determination unit 3003, configured to perform region growing based on the seed points to determine second coronary artery trunk information and second coronary artery branch information.

Fig. 18 is a schematic structural diagram of a coronary artery segmentation apparatus according to another exemplary embodiment of the present application. The embodiment shown in fig. 18 of the present application is extended based on the embodiment shown in fig. 15 of the present application, and the differences between the embodiment shown in fig. 18 and the embodiment shown in fig. 15 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 18, the coronary artery segmentation apparatus provided in the embodiment of the present application further includes:

a second post-processing operation module 500, configured to perform a second post-processing operation on the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information to correct the first coronary artery trunk segmentation information and the first coronary artery branch segmentation information, so as to obtain second coronary artery trunk segmentation information and second coronary artery branch segmentation information.

Fig. 19 is a schematic structural diagram of a second post-processing operation module according to an exemplary embodiment of the present application. The embodiment shown in fig. 19 of the present application is extended based on the embodiment shown in fig. 18 of the present application, and the differences between the embodiment shown in fig. 19 and the embodiment shown in fig. 18 will be emphasized below, and the descriptions of the same parts will not be repeated.

As shown in fig. 19, in the coronary artery segmentation apparatus provided in the embodiment of the present application, the second post-processing operation module includes:

a maximum connected domain determining unit 5001 configured to determine a maximum connected domain of a coronary artery trunk based on the first coronary artery trunk segmentation information;

a second coronary artery trunk segmentation information determining unit 5002, configured to perform region growth based on the maximum connected domain of the coronary artery trunk to determine second coronary artery trunk segmentation information;

a coronary branch connected domain determining unit 5003 for determining a coronary branch connected domain based on the first coronary branch segmentation information;

the second coronary branch segmentation information determination module 5004 determines second coronary branch segmentation information based on the coronary branch connected domain.

Fig. 20 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application. The embodiment shown in fig. 20 of the present application is extended based on the embodiment shown in fig. 15 of the present application, and the differences between the embodiment shown in fig. 20 and the embodiment shown in fig. 15 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 20, in the coronary artery segmentation apparatus provided in the embodiment of the present application, the first determination module includes:

the first coronary artery trunk branch information determining unit 1004 is configured to input aorta segmentation information, heart segmentation information, and coronary artery segmentation information corresponding to the coronary artery image to be segmented into the trunk branch model, so as to obtain first coronary artery trunk information and first coronary artery branch information.

Fig. 21 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application. The embodiment shown in fig. 21 is extended from the embodiment shown in fig. 20, and the differences between the embodiment shown in fig. 21 and the embodiment shown in fig. 20 will be emphasized below, and the descriptions of the same parts will not be repeated.

As shown in fig. 21, the coronary artery segmentation apparatus provided in the embodiment of the present application further includes:

a third determining module 1000, configured to determine a training data set and aorta segmentation information, heart segmentation information, and coronary artery segmentation information corresponding to the training data set;

a fourth determining module 1001, configured to determine external coronary artery graph information corresponding to the coronary artery segmentation information corresponding to the training data set;

a fifth determining module 1002, configured to sequentially determine external aortic graphical information, external cardiac graphical information, and external coronary graphical information based on the external coronary graphical information;

the trunk branch model generation module 1003 establishes an initial network model, and trains the initial network model based on the training data set, the aorta external graph information, the heart external graph information, and the coronary artery external graph information to generate a trunk branch model.

Fig. 22 is a schematic structural diagram of a coronary artery segmentation apparatus according to still another exemplary embodiment of the present application. The embodiment shown in fig. 22 of the present application is extended based on the embodiment shown in fig. 15 of the present application, and the differences between the embodiment shown in fig. 22 and the embodiment shown in fig. 15 are emphasized below, and the descriptions of the same parts are omitted.

As shown in fig. 22, in the coronary artery segmentation apparatus provided in the embodiment of the present application, the second determination module includes:

the first coronary artery trunk and branch segmentation information determining unit 2001 is configured to input aorta segmentation information, heart segmentation information, coronary artery segmentation information, first coronary artery trunk information, and first coronary artery branch information corresponding to the coronary artery image to be segmented into a trunk branch segmentation model, so as to obtain first coronary artery trunk segmentation information and first coronary artery branch segmentation information.

Exemplary electronic device

Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 23. Fig. 23 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application.

As shown in fig. 23, the electronic device 60 includes one or more processors 601 and memory 602.

Processor 601 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in electronic device 60 to perform desired functions.

Memory 602 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, Random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, Read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium and executed by the processor 601 to implement the coronary segmentation methods of the various embodiments of the application described above and/or other desired functions. Various contents such as a coronary image to be segmented may also be stored in the computer-readable storage medium.

In one example, the electronic device 60 may further include: an input device 603 and an output device 604, which are interconnected by a bus system and/or other form of connection mechanism (not shown).

The input device 603 may include, for example, a keyboard, a mouse, and the like.

The output device 604 may output various information to the outside, including the determined first coronary artery trunk segment information and the first coronary artery branch segment information, and the like. The output devices 604 may include, for example, a display, speakers, a printer, and a communication network and remote output devices connected thereto, among others.

Of course, for the sake of simplicity, only some of the components related to the present application in the electronic device 60 are shown in fig. 18, and components such as a bus, an input/output interface, and the like are omitted. In addition, the electronic device 60 may include any other suitable components depending on the particular application.

In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform the steps in the coronary segmentation method according to various embodiments of the present application described above in this specification.

The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.

Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the steps in the coronary segmentation method according to various embodiments of the present application described above in this specification.

The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.

The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".

It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.

The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.